Deep Automatic Natural Image Matting
This work addresses the challenge of extending image matting to more varied and complex natural images, which is useful for image editing applications, though it is incremental in improving generalization over prior methods.
The paper tackles the problem of automatic image matting for diverse natural images, including those with transparent or non-salient foregrounds, by proposing a novel network that predicts a generalized trimap and uses attention mechanisms, and it outperforms existing methods on a new benchmark dataset AIM-500.
Automatic image matting (AIM) refers to estimating the soft foreground from an arbitrary natural image without any auxiliary input like trimap, which is useful for image editing. Prior methods try to learn semantic features to aid the matting process while being limited to images with salient opaque foregrounds such as humans and animals. In this paper, we investigate the difficulties when extending them to natural images with salient transparent/meticulous foregrounds or non-salient foregrounds. To address the problem, a novel end-to-end matting network is proposed, which can predict a generalized trimap for any image of the above types as a unified semantic representation. Simultaneously, the learned semantic features guide the matting network to focus on the transition areas via an attention mechanism. We also construct a test set AIM-500 that contains 500 diverse natural images covering all types along with manually labeled alpha mattes, making it feasible to benchmark the generalization ability of AIM models. Results of the experiments demonstrate that our network trained on available composite matting datasets outperforms existing methods both objectively and subjectively. The source code and dataset are available at https://github.com/JizhiziLi/AIM.